mcp-ragchat

Provides RAG-powered AI chat for your sites by building a local vector store from your content and serving a chat widget locally.

javascript

0

GitHub Stars

javascript

Language

3 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "gogabrielordonez-mcp-ragchat": {
      "command": "node",
      "args": [
        "/absolute/path/to/mcp-ragchat/dist/mcp-server.js"
      ],
      "env": {
        "OPENAI_API_KEY": "YOUR_API_KEY"
      }
    }
  }
}

You can deploy the mcp-ragchat server to add RAG-powered AI chat to your website. It builds a local knowledge base from your content, serves a chat widget locally, and runs entirely on your infrastructure with your API key.

How to use

You interact with this MCP server through an MCP client. After you set up the server, you can seed content, start a local chat server, and generate a self-contained embed widget. Use the client to initiate setup for a domain, then test the chat, start the server, and embed the widget on your site.

Typical workflow: seed content from your site or documents, run a local server, and paste the generated script tag into your web pages to enable live AI chat with no cloud dependencies.

How to install

Prerequisites: you need Node.js and npm installed on your machine.

Clone the project.
Install dependencies and build.

git clone https://github.com/gogabrielordonez/mcp-ragchat
cd mcp-ragchat
npm install && npm run build

Configure the MCP client to run the server locally. Create a configuration under your Claude Code setup to point to the Node server and provide your API key.

Configuration and start flow

This setup runs a local Node.js server that serves the chat functionality and requires an API key for the LLM provider.

{
  "mcpServers": {
    "ragchat": {
      "command": "node",
      "args": ["/absolute/path/to/mcp-ragchat/dist/mcp-server.js"],
      "env": {
        "OPENAI_API_KEY": "sk-..."
      }
    }
  }
}

Save this as the MCP configuration (for example, under ~/.claude/mcp.json) and reference it from your MCP client. Start by seed- ing your content, then test the chat, then generate the widget embed code.

Embed the chat widget

After you seed content and run the server, you can generate a self-contained embed snippet. Paste the snippet into your website to display a floating chat bubble.

<script src="http://localhost:3456/widget.js"></script>

Available tools

ragchat_setup

Seed a knowledge base from markdown content. Each section becomes a searchable document with embeddings.

ragchat_test

Send a test query to verify RAG retrieval and LLM response quality.

ragchat_serve

Start a local HTTP chat server with CORS and input sanitization.

ragchat_widget

Generate a self-contained embed tag for a floating chat widget.

ragchat_status

List all configured domains with their document counts and configuration details.